JWiki: The AI Generated Jmail Encyclopedia for Exploring the Epstein Files

PostedAugust 5, 2026

UpdatedAugust 5, 2026

ByEpsteinWiki

JWiki, officially titled The Jmail Encyclopedia, converts information from publicly released Epstein records into searchable encyclopedia style profiles of people, companies, properties, institutions, events, and alleged criminal activity.

The tool is part of the larger Jmail project created by technologists Riley Walz and Luke Igel. Jmail originally transformed released Epstein emails into a familiar inbox interface. JWiki takes the project further by using artificial intelligence to synthesize those emails and other released records into narrative profiles.

JWiki makes a vast and difficult archive easier to explore. However, its own website states that its content is generated by artificial intelligence. Researchers must therefore treat every profile as a discovery tool rather than an authoritative source.

Snapshot

Resource name: JWiki

Official title: The Jmail Encyclopedia

Website: Jmail.world/wiki

Resource type: Searchable encyclopedia and research discovery tool

Creators: Riley Walz and Luke Igel

Public launch: February 2026

Primary material: Publicly released government documents and the Epstein email archive

Technology: Artificial intelligence synthesis, optical character recognition, email extraction, entity matching, and community editing

Public access: Available without registration

Account features: Editing proposals, drafts, watchlists, recent changes, and revision histories

Related tools: Jmail, JDrive, JPhotos, JFlights, Jamazon, Jacebook, Jotify, and Jeddit

Major limitation: Articles are generated by artificial intelligence and may contain factual, chronological, attribution, and identity matching errors

What Is JWiki?

JWiki is an encyclopedia style interface built around the Epstein records collected by the larger Jmail project.

Instead of requiring a researcher to search millions of pages individually, JWiki creates profiles that summarize the records connected to a person, company, property, or subject.

Profiles may include biographical information, the number of emails associated with a name, frequent contacts, visits to Epstein properties, financial activity, notable correspondence, legal proceedings, and connections to other subjects in the archive.

The homepage provides alphabetical browsing, category browsing, keyword searching, and direct links to the underlying Jmail email collection.

The tool also contains broader topic articles about Epstein’s properties, financial relationships, flight records, recruitment methods, the 2007 nonprosecution agreement, alleged obstruction of justice, and relevant criminal statutes.

Who Created JWiki?

JWiki was created by technologists Riley Walz and Luke Igel as part of the Jmail collection of public research interfaces.

Walz is an internet artist and software developer known for building unconventional public information tools. Igel is a technologist and the chief executive of the artificial intelligence company Kino AI.

The two developers launched Jmail in November 2025 after converting released Epstein emails into an interface resembling a familiar email inbox. According to Fast Company’s account of the project, the first version of Jmail was built in a single night.

JWiki followed in February 2026. It expanded the original email interface into an encyclopedia containing profiles, topic pages, source links, article histories, and cross references.

How JWiki Works

JWiki uses artificial intelligence to examine extracted text from publicly released government documents.

The broader Jmail project used optical character recognition to convert emails and other scanned records into machine readable text. The resulting archive was organized into simulated inboxes and searchable collections.

JWiki uses that material to generate narrative summaries. The system attempts to connect names, email addresses, companies, events, properties, and related records.

Each profile can include an email count, thread count, activity period, list of frequent contacts, biographical summary, discussion of connections to Epstein, and links to relevant emails.

The site clearly states that its content is “generated by AI from publicly released government documents.”

That disclosure is essential. Artificial intelligence can organize and summarize records quickly, but it can also merge identities, misunderstand dates, confuse senders with recipients, repeat unverified allegations, and present an inference as an established fact.

Searching by Person, Place, Company, or Subject

JWiki allows users to search for people and subjects without knowing an EFTA number or the location of the original document.

The alphabetical index includes public officials, bankers, lawyers, academics, pilots, assistants, models, business executives, members of Epstein’s household, and individuals who appeared in released correspondence.

Topic pages cover institutions and subjects such as:

JPMorgan Chase
MC2 Model Management
Little Saint James
Epstein’s Palm Beach residence
The Manhattan townhouse
Zorro Ranch
Flight logs and travel records
Victim references
Recruitment patterns
The 2007 nonprosecution agreement
Obstruction of justice evidence
Epstein’s alleged coconspirators

The site also connects each profile to the corresponding emails in Jmail. This helps researchers move from an artificial intelligence summary to the material that generated it.

JWiki’s Email Counts Require Careful Interpretation

Many profiles display large email totals. The JWiki homepage lists Lesley Groff with more than 230,000 emails, Richard Kahn with more than 92,000, Karyna Shuliak with more than 48,000, and Larry Visoski with more than 38,000.

These figures should not automatically be described as direct personal exchanges with Epstein.

An email count may include messages in which the person was the sender, recipient, copied recipient, subject, quoted participant, attachment reference, or matched entity. The archive may also contain duplicate documents, repeated email chains, forwarded messages, aliases, and multiple copies of the same material.

A large count can indicate that a person is important to the archive. It does not, by itself, establish criminal knowledge, participation, intimacy, or responsibility.

Researchers should examine the actual messages before describing the nature of a relationship.

Example: The Richard Kahn Profile

The JWiki profile for Richard Kahn places Epstein’s longtime accountant among the most frequently appearing people in the email archive.

Kahn’s prominence is consistent with released records showing his involvement in wire transfers, bank communications, trusts, corporate entities, property expenses, payroll, and Epstein’s estate.

Independent evidence supports the conclusion that Kahn occupied a central financial position. An FBI interview record states that wire instructions generally came from Kahn or included him in the communications. The original document is available at EFTA00128780.

Another record documents the transfer of approximately $33.9 million from Epstein controlled JPMorgan accounts into Deutsche Bank accounts in October 2013. That evidence is available at EFTA01578684.

EpsteinWiki provides a separate evidence based examination of Richard Kahn’s work as Epstein’s accountant, financial manager, and estate coexecutor.

The JWiki profile is useful for identifying possible emails and connections. The EFTA records remain the evidence that should support published conclusions.

Example: MC2 Model Management

The JWiki article on MC2 Model Management brings together material about the modeling agency founded by Jean Luc Brunel and Jeffrey Fuller.

The profile discusses Epstein’s financial relationship with the company, its international scouting operations, corporate debt, communications with Epstein’s accountant, and allegations that the modeling network facilitated access to young women and girls.

JWiki cites a JPMorgan due diligence report stating that MC2 received $1 million from Epstein in 2005. The report questioned whether the money represented an undisclosed investment or payment connected to procuring women.

The JWiki entry also organizes emails concerning MC2 expenses, credit arrangements, scouting activity, staff, tax liens, and communications with Richard Kahn.

This is a strong example of JWiki’s value. The platform places scattered emails, bank records, media reports, and corporate information into one readable narrative.

It is also an example of why verification matters. Allegations found in bank compliance materials or court filings must remain attributed to those sources. A due diligence report recording an allegation does not automatically prove that every statement inside the report is true.

For additional evidence based context, see EpsteinWiki’s investigation of Jean Luc Brunel’s modeling network and how Epstein’s business and trafficking system worked.

Example: Virginia Giuffre’s Profile

The JWiki profile for Virginia Giuffre compiles information from court filings, flight logs, released emails, her public testimony, and her memoir.

The profile includes a detailed travel chronology and connects Giuffre’s account with entries in Epstein’s flight records.

One of the evidence files cited by JWiki is EFTA00151067, a flight log record containing entries associated with Virginia Roberts Giuffre and other passengers.

JWiki also identifies Giuffre’s sworn declaration as EFTA01139414.

The profile illustrates both the potential and the danger of automated synthesis. Gathering court filings, flight records, testimony, and later reporting into one page can help researchers understand chronology. However, survivor records require exceptional care. Artificial intelligence should not determine which allegations are credible, expose private identities, or flatten complex testimony into an automated conclusion.

Researchers should prioritize Giuffre’s own words, authenticated court records, and verified evidence.

Revision Histories and Community Editing

JWiki includes article histories, draft pages, discussion pages, watchlists, and a recent changes section.

Users can create accounts and propose edits. Available reporting indicates that proposed changes are reviewed by administrators before publication.

Revision histories allow researchers to see how an article changed over time. This is valuable when a profile contains a disputed statement, corrected identity, updated source, or altered interpretation.

The editing system makes JWiki more transparent than an artificial intelligence summary that cannot be inspected or corrected. However, administrator review does not transform an entry into a verified legal or historical record.

Researchers should record the date on which they consulted a page because its content may change.

A Documented Warning About Timeline Errors

JWiki’s Bella Klein profile demonstrates why users must verify artificial intelligence generated claims.

The profile displays an activity range extending into November 2023. It also states that Klein asked Epstein a financial question in October 2023.

Jeffrey Epstein died on August 10, 2019. He could not have received or answered an email in 2023.

The discrepancy may result from a quoted email inside a later document, incorrect metadata, a forwarded message, optical character recognition failure, thread reconstruction, or an artificial intelligence error. Regardless of the cause, the statement cannot be read literally.

This example does not make the entire resource useless. It shows that polished prose and precise looking dates can still contain obvious errors.

Every important date should be verified in the original email or evidence file.

Artificial Intelligence Can Confuse Mentions With Relationships

A person may appear in an email without communicating with Epstein.

Their name might occur in a news clipping, address list, forwarded article, calendar entry, legal document, quoted message, attachment, or conversation between other people.

Artificial intelligence may describe such an appearance as correspondence or a relationship even when the underlying record shows only a third party reference.

The same problem applies to identity matching. People with similar names may be merged. One person may appear under several spellings or email addresses. A shortened name may be assigned to the wrong individual.

Researchers must confirm:

Who sent the message
Who received it
Who was copied
Whether the relevant language came from the sender or from quoted material
Whether the document is an original email or a later reproduction
Whether the named individual is correctly identified
Whether the same message appears multiple times
Whether the date reflects the original communication or a later archive event

A Name in JWiki Does Not Establish Wrongdoing

JWiki contains profiles of people with very different relationships to the Epstein record.

Some were employees or financial administrators. Some were social acquaintances. Some were journalists, academics, lawyers, pilots, politicians, models, or business contacts. Some were survivors. Others appeared only because their names were mentioned in correspondence.

Inclusion in JWiki does not establish that a person committed, witnessed, facilitated, or knew about a crime.

An email count does not establish guilt. A photograph does not establish criminal participation. A calendar entry does not prove that a meeting occurred. A flight log records reported travel, but it does not establish what happened at the destination.

Responsible research requires clear distinctions among association, documented communication, allegation, corroborated conduct, civil findings, criminal charges, and criminal convictions.

JWiki’s Strongest Features

JWiki makes a massive archive accessible to people who do not have specialized document analysis skills.

Its strongest features include searchable profiles, direct links to Jmail emails, alphabetical browsing, topic pages, cross references, email counts, frequent contact lists, revision histories, and community correction tools.

The familiar encyclopedia format lowers the barrier to entry. A researcher can begin with a person or institution and quickly identify related properties, companies, associates, and correspondence.

JWiki is especially useful for generating search terms. A profile may reveal an unfamiliar company name, email address, assistant, property, date, or financial entity that can then be searched in Epstein Data.

JWiki’s Most Important Limitations

JWiki is not an authenticated evidence archive.

Its articles are generated by artificial intelligence. They may include hallucinated conclusions, mistaken identities, inflated email totals, duplicated records, incorrect dates, unsupported legal analysis, or language stronger than the underlying evidence permits.

The interface can also create an authority problem. Because the pages resemble Wikipedia, readers may assume that they underwent the same editorial and sourcing process. They did not.

The site sometimes discusses possible criminal statutes. Researchers should not repeat those sections as legal findings. Only courts and appropriate legal authorities can determine whether a person violated a criminal law.

JWiki should never be the only source cited for a serious allegation.

How Researchers Should Use JWiki

Begin by searching JWiki for a person, company, institution, property, or event.

Read the profile to identify relevant names, dates, entities, and themes. Then select the link to view the associated emails in Jmail.

Open each important email and determine its sender, recipients, original date, full thread, attachments, and context.

Search the corresponding EFTA number in Epstein Data whenever one is available.

Compare the claim with court filings, government records, corporate registrations, authenticated flight logs, sworn testimony, and reliable reporting.

When publishing, cite the original evidence. Credit JWiki when it materially helped locate or organize the records.

Relationship to EpsteinWiki and Epstein Data

JWiki, EpsteinWiki, and Epstein Data perform different research functions.

JWiki uses artificial intelligence to generate profiles and suggest connections.

Epstein Data provides direct access to evidence files and searchable records.

EpsteinWiki builds documented explanatory articles that distinguish evidence, allegations, findings, and unresolved questions.

A responsible workflow begins with discovery, moves to evidence, and ends with contextual analysis.

JWiki can help researchers discover a lead. Epstein Data can verify the underlying record. EpsteinWiki can explain the broader meaning of that evidence without treating automated output as proof.

Key Takeaways

JWiki is an artificial intelligence generated encyclopedia built from publicly released Epstein records.
Riley Walz and Luke Igel created the tool as part of the larger Jmail research project.
JWiki organizes people, companies, properties, events, emails, travel records, and major investigative subjects.
Profiles link users to associated emails in Jmail and include revision histories and community editing tools.
Email totals may include mentions, copied recipients, quoted threads, aliases, and duplicate records. They should not automatically be described as direct correspondence.
JWiki can contain impossible dates and other artificial intelligence errors, as demonstrated by the Bella Klein profile’s reference to communication with Epstein in 2023.
A person’s inclusion in JWiki does not establish knowledge of or participation in Epstein’s crimes.
Serious claims must be verified through original EFTA files, court records, government documents, or other primary evidence.
JWiki is most valuable as a discovery and navigation tool.
JWiki should never be cited as the sole evidence for an allegation of criminal conduct.

Why JWiki Matters

The Epstein releases created an enormous accessibility problem. Millions of pages may technically be public while remaining functionally invisible to anyone who cannot download, search, organize, and cross reference them.

JWiki attempts to solve that problem by turning the archive into a familiar encyclopedia. It can expose patterns, locate overlooked correspondence, and help researchers understand how people, companies, properties, and financial systems intersect.

Its speed and accessibility are genuine strengths. Its artificial intelligence foundation is also its central weakness.

JWiki is best understood as a powerful research compass. It can point toward important material, but it cannot replace the original evidence.

Sources

Tags:

Uncategorized